MIT Team Isolates the Stratospheric Fingerprints of Pinatubo, Australian Wildfires, and Hunga Tonga
The atmosphere is never quiet. At any moment, dozens of processes are pushing global temperatures up or down - ocean heat exchange, aerosol loading from industrial emissions, shifts in cloud cover, the slow oscillation of El Nino and La Nina. Against that backdrop of continuous variation, detecting the specific temperature signature of a single event - a volcano, a wildfire season - is like picking out one instrument in a full orchestra playing fortissimo.
MIT scientists have developed a method to quiet enough of that background noise to extract the specific atmospheric signals of major natural events, and applied it to three cases: the eruption of Mount Pinatubo in the Philippines in 1991, the catastrophic Australian bushfire season of 2019-2020, and the underwater eruption of Hunga Tonga in the South Pacific in January 2022. All three events produced statistically significant and measurable changes in global stratospheric temperatures. The study appears in the Proceedings of the National Academy of Sciences.
Why stratospheric detection is hard
Eruptions and large fires inject material into the stratosphere - sulfur dioxide from volcanoes forms sulfate aerosols that reflect sunlight and cool the surface; soot and organic carbon from wildfires absorb solar radiation and heat the stratosphere while potentially cooling the surface below. These effects are real and have been estimated from climate models, but confirming them directly from observational data requires separating the event signal from simultaneous variability in ocean temperatures, solar output, and the natural variability of atmospheric circulation patterns.
Previous analyses of the Pinatubo eruption, for instance, confirmed broad surface cooling effects, but extracting a clean signal has required careful statistical treatment and has been subject to debate about the magnitude and duration of the effect. The Australian fires and Hunga Tonga eruption are more recent and thus less studied, but both were unusual enough in character to raise questions about their atmospheric footprint.
Three events, three distinct signals
The MIT team's approach uses a detection framework designed to distinguish event-driven changes from background variability. Applied to global atmospheric temperature data, the method found that all three events left identifiable marks specifically in the stratosphere - the atmospheric layer sitting above the troposphere and below the mesosphere, roughly 12 to 50 kilometers above the surface.
Mount Pinatubo's 1991 eruption injected roughly 20 million metric tons of sulfur dioxide into the stratosphere, the largest volcanic injection since Krakatoa. Global average surface temperatures fell by approximately 0.5 degrees Celsius in the following year. The MIT analysis found the stratospheric signal consistent with these known effects.
The 2019-2020 Australian fires burned approximately 18.6 million hectares, releasing a estimated 715 million metric tons of carbon dioxide equivalent along with large quantities of smoke and aerosols. The fires were notable for producing a pyrocumulonimbus events - fire-driven thunderstorms that can inject smoke directly into the stratosphere, far above the altitude that typical fire plumes reach. The detection of a stratospheric temperature signal from this event supports the interpretation that the Australian fires had climatic effects extending well above the boundary layer where most fire emissions remain.
Hunga Tonga's January 2022 eruption was unusual in a different way. It injected an estimated 146 teragrams of water vapor into the stratosphere - far more than typical volcanic eruptions, which tend to be dominated by sulfur dioxide. Water vapor is a potent greenhouse gas, and its injection at stratospheric altitudes raised concerns about warming effects that could persist for years as the water vapor gradually dispersed. The MIT analysis detected a stratospheric temperature signal from this event as well, though the character of the warming effect differs from the cooling typically associated with sulfate aerosol injection.
Implications for climate attribution
The ability to detect individual event signals matters for climate attribution - the effort to assign specific causes to specific observed changes. As climate science becomes more sophisticated and the demands on it from policy and litigation contexts increase, the capacity to say with statistical confidence that a particular event produced a particular atmospheric response becomes more valuable.
The method developed by the MIT team represents a contribution to that capacity. It does not, by itself, resolve debates about the magnitude of volcanic or wildfire climate effects, which depend on additional factors including aerosol optical properties, atmospheric circulation responses, and feedbacks not captured in the temperature signal alone. But demonstrating that the signals are detectable in observational data, rather than only visible in model output, strengthens the empirical foundation for those estimates.
One area of ongoing interest is whether the Hunga Tonga water vapor injection has contributed to observed warming in the years since the 2022 eruption, a question that the stratospheric temperature signal alone cannot fully resolve but that the MIT work helps to frame.